Magnetic Steering and ECCD Mitigation of locked NTMs

1
Magnetic Steering and ECCD Mitigation of locked
NTMs
Presented byFrancesco Volpe In collaboration
with R.J. Buttery, R.J. La Haye,
G. Jackson, M. Okabayashi, F.W. Perkins, R.
Prater, H. Takahashi Acknowledgements to J.
Ferron, A. Hyatt, B. Johnson, B. Penaflor, C.
Petty, H. Reimerdes, E.J. Strait ECH
Team Presented at the Friday Science Meeting,
GA September 22, 2006
2
Motivation

• ITER expected to have slow (few kHz) NTMs ? high
  probability of locking.
• Locked 2/1 modes are a recurrent cause of
  disruptions.
• ECCD proved successful in suppressing rotating
  NTMs.
• Success of ECCD for locked modes not guaranteed,
  due to island locking in a position not
  accessible by gyrotrons.
• Resonant Magnetic Perturbations (RMPs) from
  I-coils used at DIII-D for RWM and ELM control.
• Although more challenging (modes lie deeper in
  the plasma and are faster), here RMPs are used to
  control NTM rotation and assist their ECCD
  Stabilization

3
Overview of EFCECCD control proposals
B1 B2 B3 B4
4
How to predict a Locked Mode?

• 3 types of dud detectors were used
• dB/dt gt20T/s for 200ms
• Mode frequency drops below 1kHz
• Born-locked mode gt5G for 20ms

B(G) f(kHz)
dB/dt (T/s)
5
Exp B1 Controlled locking (accessible by
gyrotrons). Then apply cw ECCD.
Some beams inherited from mode onset phase
(changing from shot to shot).
Modify beams (via J.Ferrons algorithm) to get
low plasma and mode rotation, unless mode slows
down for free
NBI
ctr
co
Dud detector detects oncoming locking
f21
Due to overcorrection, mode locks even faster,
but to toroidal phase optimal for CW ECCD
1kHz
Generate a static error field with I-coils Ratio
between IU30, IU90 and IU150 experimentally found
(via phase scan) to lock O-point in front of
gyrotron. If time, optimize strength of
magn.perturbation by changing all currents in
proportion
DC currents in I-coils
Max ECCD
Other shots inject in X-point and intermediate
phase angles for comparison
6
Toroidal Alignment at locking, 0.66Hz I-coil
travelling wave and CW ECH were applied change
in mode amplitude observed.
7
CW ECCD EFC plasmaMode amplitude is larger
and varies (4-7.5G) with I-coil frequency
8
A Vacuum shot was taken to subtract I-coils and
C-coils effect on magnetics. Saddle loops measure
1G, constant.
9
Even after subtraction of Vacuum Shot, locked
mode changes amplitude when toroidally steered
and illuminated by ECH
Non-uniform rotation consistent with mode
amplitude and stronger or weaker wall braking
10
Vacuum shot
11
No-ECH Ref. Shot has potential for isolating
interaction between mode intrinsic error, but
it also affects ne, b, etc. and makes island more
sticky
12
ECH deliberately misaligned by 3 lower Ip and
BT same heating, same mode-error field
interaction but no ECCD stabilization.
Subtraction made tricky by these instabilities
Exp. can be improved by 22 gyrotrons or real
time ECH steering
13
Clear Difference in Phase
Consistent with the fact that two different
phases are relevant Between island and ECCD in
ECH case (top) Between I-coil and Saddle-Loops
in no-ECH case (bottom)
14
ECH aligned to island via radial jog of plasma.
Improved stabilization for DR-2cm. When ECH off,
mode amplitude rises again.
15
Exp. B2 Prevent locking by sustaining mode
rotation. Then apply ECCD (modulated or cw).
Like B1
NBI
ctr
co
Dud detector detects oncoming locking
f21
Rotation at 1kHz sustained by dynamic EFC
1kHz
Generate a dynamic error field by AC currents of
f1kHz. Repeat for various amplitudes at fixed
ratios between pairs of coils.
Amplitude of AC currents in I-coils
ECCD modulated at 1kHz

• Shotplan
• No ECCD (for ref.)
• MECCD at 1.001kHz (equivalent to phase scan, if
  I-coils at 1kHz)
• MECCD at 1kHz in O-point (and in another shot in
  X-point, for comparison)
• CW ECCD for reference

16
Exp. B3 Spin mode up. No ECCD necessary.Exp. B4
is like B3 but wait for locking first, then
unlock spin up.
Find co/ctr mix such that mode rotates at 1kHz
more or less constantly. No slowing down here.
Alternatively, sustain rotation at 1kHz with
I-coils (see B2), then apply steps in I-coil
w/forms.
NBI
ctr
co
Apply EFC rotating at ff21 to control mode
rotation
3
f I-coils
f21
2
Apply EFC accelerating from fltf21 to f f21 to
make mode spin faster and faster. Invert ramp to
study w12-b histeresis.
1
1 2 3
w21
Modulation of I-coil not necessarily commenced
here might have pre-existed, if rotation was
magnetically sustained (see fig.2)
3 2 1
b
Renounced to f/back on I-coil, not ready.
17
Gradual Entraining of I-coil travelling wave,
1-100Hz
18
Gradual Entraining of I-coil travelling wave,
1-100Hz
19
Gradual Entraining of I-coil travelling wave,
1-100Hz
20
Travelling Wave of up to 60Hz coupled to 2/1 mode
21
I-coil Travelling Wave less effective at high
frequencies

• Current II-coil delivered by SPAs falls off with
  f
• Besides, for the same II-coil, the magnetic
  perturbation exerted on the plasma decreases due
  to
• Partial compensation from image currents in the
  wall (skin effect? Distance between I-coil
  image comparable with wavelength?) -3dB _at_100Hz,
  -10dB _at_1kHz.
• More Shielding associated with (faster) rotation
• Furthermore, the same BI-coil couples less
  effectively with a faster, rotationally
  mitigated, weaker mode ( compass of reduced
  sensitivity immersed in the same field)
• Phase delays in SPAs
• SPAsSwitching Power Amplifiers ? discrete steps

22
Summary of Experiments carried out

• Preferential Locking (CW ECCD DC I-coils)
• CW ECCD 0.66Hz I-coil Travelling Wave
• Toroidal alignment of island relative to
  gyrotrons
• Radial Jog of Plasma
• Radial alignment
• CW ECCD DC I-coils
• Injection in O-point and, for comparison, in
  X-point.
• Sustained Rotation (MECCD AC I-coils)
• Ramps of I-coil Frequency